Sand Control Measures and Sand Drift Fences

Sand Control Measures and Sand Drift FencesAbdulmalik A. A. Alghamdi1 and Nasser S. Al-Kahtani2

Abstract: Sand drift and sand dune movements are typical logistic problems facing civilian and industrial cities in arid and semiaridcountries like Saudi Arabia. Some of these countries are considered active when it comes to sand drift and sand dune movement, due tothe high annual sand drift rate. Urban cities have extensive facilities in the middle of these active areas that require good protection andinnovative solutions to this problem. This paper briefly reviews sand movement control measures and highlights sand drift fence designguidelines for the first time.

DOI: 10.1061/�ASCE�0887-3828�2005�19:4�295�

CE Database subject headings: Sand; Drift; Fences; Dunes; Barriers; Deserts; Saudi Arabia.

Introduction

Approximately one-third of the Arabian Peninsula is covered bysand dune areas. The Empty Quarter �Rub Al-Khali� probablycontains the largest continuous sand dune area in the world, cov-ering an area of nearly 600,000 km2. Another 180,000 km2 iscovered by the Great Nafud and Ad Dahna deserts in the interiorpart of the peninsula �Fig. 1�. The Jafura sand sea covers theArabian Gulf coastal region in the Eastern Province of SaudiArabia. The Jafura sand sea extends from an area of high windenergy in the north to an area of low wind energy in the south,resulting in a general sand drift direction southwards to southeast-wards. Fryberger et al. �1983� zoned the eolian landscape of theJafura sand sea into an area of deflation �erosion� on the north, anarea of deposition in the south, and an area of transport in be-tween. Most of the relevant literature is focused on sand dunemodeling and advancement and little is revealed regarding sanddrift fencing �see, e.g., Qong �2000�; Amarouchene et al. �2001�;Lima et al. �2002�; and Kroy et al. �2002��.

The Jafura sand sea of the Eastern Province is an elongated32–250 km wide sand body extending along the western side ofArabian Gulf from Kuwait in the north to the Empty Quarter inthe south, a distance of about 800 km �Research Institute 1985�.This body forms narrow band in the north then widens in thesouth where it merges with Empty Quarter. The annual sand driftrate varies from 10 to 120 tons/m in the Jafura sand sea andcan reach higher values in very active areas �Bagnold 1971;Dhir 1995�.

As a result of this, sand control is a pressing necessity to

1Associate Professor, Dept. of Mechanical Engineering, KingAbdulaziz Univ., P.O. Box 80299, Jeddah 21589, Saudi Arabia. E-mail:[email protected]

2Geotechnical Engineer, Civil Engineering Unit, Mechanical andCivil Engineering Division, Consulting Services Dept., Saudi Aramco,Dhahran 31311, Saudi Arabia. E-mail: [email protected]

Note. Discussion open until April 1, 2006. Separate discussions mustbe submitted for individual papers. To extend the closing date by onemonth, a written request must be filed with the ASCE Managing Editor.The manuscript for this paper was submitted for review and possiblepublication on September 3, 2003; approved on May 11, 2004. This paperis part of the Journal of Performance of Constructed Facilities, Vol. 19,No. 4, November 1, 2005. ©ASCE, ISSN 0887-3828/2005/4-295–299/
$25.00.
JOURNAL OF PERFORMANCE

minimize hazardous impacts on existing facilities. National gov-ernments spend tens of millions of U.S. dollars for contractorsremoving accumulated sand near civilian and industrial facilities.For example, Saudi Aramco �the Saudi Arabian oil producingcompany� uses 300,000 barrels of heavy crude oil �tar-oil� peryear for sand dune stabilization along pipelines, roads, runways,and foundations. Some of the Saudi Aramco facilities that expe-rience these problems are gas-oil separation plants, pump stationsof the East-West Pipeline, and skid roads. Saudi Aramco has rec-ognized the importance of sand control measures since the start ofits operation �Kerr and Nigra 1952�. The latest study ofsuch controls was completed in 1998 by the Research Institute atKing Fahd University of Petroleum and Minerals �KFUPM��Research Institute 1998�.

Generally speaking, worldwide sand control measures havebeen applied based on various studies and on personal judgmentthat lacks scientific and engineering consideration. Thus, in someplaces, ill-designed measures actually create sand hazards ratherthan controlling them. This paper outlines sand movement controlmeasures and discusses proper sand fence design as an effectivemeans of sand movement control.

Fig. 1. One-third of Arabian Peninsula is covered by sand dunes

OF CONSTRUCTED FACILITIES © ASCE / NOVEMBER 2005 / 295

Sand Movement and Its Impact on ExistingFacilities

Sand Transport

Sand is transported by three different mechanisms. Suspensionaccounts for 5% of the transport, involving very small particlesizes ��100 �m�. Almost 75% of the sand travels by saltation, atrajectory movement for medium sized particles. Large particlesmove by surface creep and represent 20% of the transportedsands. The details of sand transport, sand deposition mechanisms,and sand dune formation can be found in Bagnold �1971�.

Sand movement is highly affected by geomorphology �vegeta-tion, shapes and height of terrains, and grain sizes of the sand�and wind energy. The most important factors that affect sandmovement are summarized as follows:1. Wind speed: A minimum of 6–8 m/s wind speed is needed

for sand transport.2. Sand particle size: Heavy particles creep on the surface,

while small ones fly in suspension.3. Terrain type: Sand transport is rare on sabkha �sabkha is a

sand sheet with near-surface seawater� surface �see Fig. 2�.The reason for this is not very clear, but one possible sce-nario attributes this to the wet surface of the sabkha.

Fig. 2. Sand dune advancement is limited on sabkha surfaces

Fig. 3. Vegetation helps to stabilize sand dunes

296 / JOURNAL OF PERFORMANCE OF CONSTRUCTED FACILITIES © AS

4. Vegetation cover: This increases surface roughness, thus de-creasing surface wind speed �Fig. 3�.

5. Groundwater: Areas with shallow groundwater resist sanddune advancement.

6. Precipitation: This directly affects surface hardness and veg-etation.

7. Temperature: High temperatures increase the mobility ofsand particles.

Impact on Existing Facilities

Many government and private sector facilities are located in sandsea areas. Sand drifts and sand erosion dictate budgets to combatthese problems, mainly mechanical removal, crude oil splashing,and construction of sand barriers. In general, infrastructures—whether roads, pipelines, industrial facilities, airports, or indus-trial areas—are subjected to sand drift accumulation and winderosion that undermine the facility and require continuous main-tenance and budgeting costs to perform temporary remedial solu-tion. The following is a list of some of the typical problems thatface onshore facilities:1. Erosion of pipeline berms and anchors �Fig. 4�;2. Road blockage by sand drift �Fig. 5�;3. Wind erosion of utilities foundations �Fig. 6�;4. Sand dunes forming and surrounding facilities �Fig. 7�; and

Fig. 4. Wind erosion of pipeline berms

Fig. 5. Sand drift into right of way or access road

CE / NOVEMBER 2005

5. Megabarchan sand dune advancement to facilities �Fig. 7�.Sand dunes can take many shapes, but the most commonones are barchans and crescents.

Sand Control Measures

The importance of sand control was recognized from the begin-ning of Saudi Aramco operations in the field. Attempted solutionsto this problem have included sand fences, sand traps, diversionwalls, and barriers. However, these measures were found to beinsufficient in some situations and require the use of bulldozerand dump trucks to remove sand accumulated near the facility.Common measures are reviewed briefly hereafter.

Vegetation

Vegetation increases the surface roughness of the sand sheet orsand dune surface and thus stops sand migration. Although thismeasure is an environmentally sound solution, it is impracticaland expensive when it comes to open desert with low precipita-tion rates. In addition to this, loose particles of sand have lowwater holding capacity and low plant nutrients. Vegetation mightbe considered as a valuable solution in the vicinities of commu-nity areas, where treated sewage water can be used for irrigationof vegetation grass and bushes.

Fig. 6. Erosion of power poles located in mobile sand

Fig. 7. Shaybah community surrounded by mega sand dunes; com-munity is well protected using variety of protection systems, includ-ing system of sand drift fences


Sand Stabilization

Sand stabilization is achieved by covering a thin layer of sanddune with a chemical stabilizer. This treatment will result in abrittle crust surface that can be broken easily due to man, animal,or vehicle mobilization. A very important requirement in the sta-bilizer is that it must be a vegetation supportive and environmen-tally acceptable substance. As mentioned previously, SaudiAramco used about 300,000 bbl/year of heavy crude oil �tar-oil�for general purposes including sand control and berm stabiliza-tion. Oil stabilization was found to be effective in large areas andwas recognized as the most effective solution �Asi et al. 2002�.However, its disadvantage is the appearance and its tendency tosoil vehicles, animals, and man who must travel though it.

Other sand control measures include:1. Ditching: digging a cut perpendicular to sand drift direction

�a very expensive solution�.2. Trenching: dune destruction using bulldozers.3. Fencing: which will be discussed in the following section.4. Sand removal at a bulk removal rate of 0.5 US$/m3 and a

manual rate of 2 US$/m3. However, improper sand removalto the wrong destination will result in sand recycling. Also, itis recommended to designate hazard sites as constructionmaterials supply sites �Fig. 8�.

5. Enhancing sand transport by terrain smoothing.

Fig. 8. Contractor removing sand using dump truck

Fig. 9. Front view of suggested sand drift fence made of carbon steelposts, wooden slats, and tension wires


Sand Drift Fences

Sand drift fences are used to deposit sands in their vicinity. Thefence is installed perpendicular to the wind direction. The direc-tion before the fence is called the backward direction while thedirection in front of the fence is called the forward direction. Themajority of sand is captured in the forward wind direction. Sanddeposition takes place due to the sudden drop in the airstreampressure in the neighborhood of the fence �Bofah and Owusu1986�.

Type of Sand Fences

There are several types of sand fences available in the industry.These are:1. Vertical slat fence, shown in Figs. 9–13;2. Horizontal slat fence �sometimes called a snow fence�;3. Palm-leaf fence �also known as a brush or Arabian fence�,

which is usually similar to the two preceding fences exceptthat palm leaves are used instead of slats; and

4. Jet fence �or European fence�. This fence is a vertical-slatfence but with tapered slats.

Investigations �Research Institute 1985, 1998� have revealed thatthe vertical slat fence is the best fence when it comes to simplic-ity, ease of manufacturing and erection, performance, and main-tenance. Thus, only the vertical slat fence is discussed here.

Fig. 10. Corrosion of low carbon steel post of sand drift fence

Fig. 11. Improper location caused collapse of this fence

298 / JOURNAL OF PERFORMANCE OF CONSTRUCTED FACILITIES © AS

Sand Fences Design Guidelines

Based on previous studies on sand drift fences �Research Institute1985, 1998; Kerr and Nigra 1952; Aramco 1984�; sand fencedesign guidelines may be summarized as follows:1. The sand fence must be perpendicular to the prevailing wind

direction.2. The optimum porosity is found in the range of 40–60% �po-

rosity ratio is the ratio of the space between slates to the slatarea�.

3. As the porosity increases to a maximum value, frontal accu-mulation decreases and backward accumulation increases.

4. A V-shape fence is not recommended �V-shape in the eleva-tion projection�.

5. Avoid using plastic slats because of their brittle behavior dueto sun exposure.

6. A vertical slat fence has better aerodynamic stability whencompared to a horizontal one.

7. Efficiency increases with the decrease in slat width.8. A solid diversion fence is not recommended.9. A solid fence with zero porosity has poor performance.10. A vertical fence tends to scour posts.11. The recommended fence height �H� is 2 m.12. A sand fence can accumulate 10H2 m3 per meter length.13. Erection of an extended fence is better than adding a new

parallel fence.

Fig. 12. Shallow depth of steel posts caused in-place toppling offence

Fig. 13. Animals attack fence if it blocks their right of way

CE / NOVEMBER 2005

14. The recommended aspect ratio length/height is 40.15. Allow for ten H as a boundary condition �end effect�. This

means for a facility of 50H width, a fence with 70H width isrequired.

16. Avoid fence leaning because it makes it difficult to extend anew fence on top of the leaning fence. Leaning means tiltingthe fence, usually in the backward direction.

17. Distance to the facility should be at least 100H.18. If a gap is needed between two parallel fences, the overlap

length must be at least 10H.19. The suggested bottom gap is 10% of the height, to avoid

direct drift accumulation.20. A palm-leaf fence �Arabian fence� is efficient and easy to

install.21. Parallel fences may be used for a high rate of drift.22. The recommended spacing between parallel fences is 40H.23. The slat material should be wood treated against insects,

1.8 m by 4 cm by 1 cm.24. Posts should be made of carbon steel pipes, 50 mm in diam-

eter, 3 m long, 3 m apart, and 4 mm in thickness, paintedwith primary and protective coating.

25. Galvanized or coated wires should be used, gauge 10 orhigher.

The preceding design parameters are valuable for specification,contracting, and cost estimation. A detailed sketch of the recom-mended sand fence is shown in Fig. 9.

Fence Life

Life of the fence and price per meter length can be calculated asfollows. Assume the sand drift rate is 10 m3/m/year, the fenceheight is 2 m, and the bulk sand removal cost is 0.5 US$ /m3.This fence can capture a volume of 40 m3 during its operationallife. Thus, the expected life is 4 years. The cost of removing40 m3 of sand is 20 US$. Thus, the breakeven point for the fenceis 20 US$ /m. In other words, a fence is the better choice if theerection cost is less than 20 US$ /m, and bulk sand removal isbest if the erection cost is more than 20 US$ /m.

Fence Maintenance

Impounding fences will eventually be buried. At that point, thesand must be removed mechanically or another fence system mustbe installed. However, mechanical removal of sand creates notonly an expense but also a problem of where to put the sandwithout creating additional hazards. Therefore, installing a newfence is usually preferred. Still, if an existing fence system is stillin good shape, it is recommended that its life be increased byadding an extension, perhaps 1 m in height, as preferred.

Sand Fence Failure Modes

Field observation of existing sand fences showed that fences arenot properly maintained. Also, it was observed that some of thesefences were not located properly. Based on field observation, sandfence failure modes are as follows:


1. Post, wire, and cable corrosion �Fig. 10�;2. Improper location of the fence �Fig. 11�;3. Post scouring due to shallow depths �Fig. 12�;4. Animal attack and human vandalism �Fig. 13�;5. Gap erosion or bad design, mainly short heights �Fig. 12�;

and6. Deterioration of plastic slats, mainly due to sun exposure.

Conclusions

Many of the public and private sector’s facilities as well as publicroads and private farms are located in the stream of sand drift inarid and semiarid countries. Sand control measures are reviewedbriefly and technical experience in design of sand drift fences hasbeen updated and summarized in this paper.

Acknowledgments

The first writer thanks King Abdulaziz University in Jeddah fortheir support during his 2001–2002 sabbatical leave. Deep appre-ciation is extended to Salih Alidi, senior engineering consultant atthe Consulting Services Department, Saudi Aramco, for his valu-able technical input into this paper.

References

Amarouchene, Y., Boudet, J. F., and Kellay, H. �2001�. “Dynamic sanddunes.” Phys. Rev. Lett., 86�19�, 4286-4289.

Aramco Engineering. Rep. AER 5241, �1984�. “Sand stabilizing and roadoil in the eastern province.” Dhahran, Saudi Arabia.

Asi, I. M., Al-Abdul Wahhab, H. I., Baghabra Al-Amoudi, O. S., Khan,M. I. and Siddiqi, Z. �2002�. “Stabilization of dune sand using foamedasphalt.” Geotech. Test. J., 25�2�, 168–176.

Bagnold, R. A. �1971�. The physics of blown sand and desert dunes,Chapman and Hall, London.

Bofah, K. K., and Owusu, Y. A. �1986�. “The eolian problems arisingfrom desertification.” Environ. Monit. Assess., 6�3�, 283–292.

Dhir, R. P. �1995�. “Problem of desertification in the arid zone of Rajas-than: A view.” Desertification Control Bulletin, 27, 45–52.

Fryberger, S. G., Al-Sari, A. M., and Clisham, T. J. �1983�. “Eolian dune,interdune, sand sheet, and siliciclastic sabkha sediments of an off-shore prograding sand sea, Dhahran area, Saudi Arabia.” AAPG Bull.,67�2�, 280–312.

Kerr, R. C., and Nigra, J. O. �1952�. “Eolian sand control.” Am. Assoc.Pet. Geol. Bull., 36�8�, 1541–1573.

Kroy, K., Sauermann, G., and Herrmann, H. J. �2002�. “Minimal modelfor sand dunes.” “Phys. Rev. Lett., 88�5�, 054301.

Lima, A. R., Sauermann, G., Herrmann, H. J., and Kroy, K.�2002�. “Modelling a dune field.” statistical mechanics and its appli-cations, 310�3-4�, 487–500.

Qong, M. �2000�. “Sand dune attributes estimated from SARimages.” Remote Sens. Environ., 74�2�, 217–228.

Research Institute. �1985�. “Blown sand control research study at Madi-nat Al-Jubail Al-Sinaiyah.” PN 23020, King Fahad Univ. of Petroleumand Minerals, Dhahran, Saudi Arabia.

Research Institute. �1998�. “Testing sand control measures with minimalenvironmental impact.” SAER 5592, King Fahad Univ. of Petroleumand Minerals, Dhahran, Saudi Arabia.


Documents

Sand Control Measures and Sand Drift Fences